The Halo Occupation Distribution of X-Ray-Bright Active Galactic Nuclei: A Comparison With Luminous Quasars

Abstract

We perform halo occupation distribution (HOD) modeling of the projected two-point correlation function (2PCF) of high-redshift (z ∼ 1.2) X-ray-bright active galactic nuclei (AGN) in the XMM-COSMOS field measured by Allevato et al. The HOD parameterization is based on low-luminosity AGN in cosmological simulations. At the median redshift of z ∼ 1.2, we derive a median mass of 1.02 +0.21 −0.23 × 10 13 h −1 M ⊙ for halos hosting central AGN and an upper limit of ∼ 10% on the AGN satellite fraction. Our modeling results indicate (at the 2.5σ level) that X-ray AGN reside in more massive halos compared to more bolometrically luminous, optically-selected quasars at similar redshift. The modeling also yields constraints on the duty cycle of the X-ray AGN, and we find that at z ∼ 1.2 the average duration of the X-ray AGN phase is two orders of magnitude longer than that of the quasar phase. Our inferred mean occupation function of X-ray AGN is similar to recent empirical measurements with a group catalog and suggests that AGN halo occupancy increases with increasing halo mass. We project the XMM-COSMOS 2PCF measurements to forecast the required survey parameters needed in future AGN clustering studies to enable higher precision HOD constraints and determinations of key physical parameters like the satellite fraction and duty cycle. We find that N 2 /A ∼ 5 × 10 6 deg −2 (with N the number of AGN in a survey area of A deg 2 ) is sufficient to constrain the HOD parameters at the 10% level, which is easily achievable by upcoming and proposed X-ray surveys.

Clustering on very small scales from a large sample of confirmed quasar pairs: does quasar clustering track from Mpc to kpc scales?

“…Hopkins et al (2007) argued instead that strong quasar clustering on small scales is simply indicative of quasars occupying group-scale or "merger-prone" environments. More recently, the small-scale clustering of quasars has been modeled using the "onehalo" term in the HOD (e.g., Kayo & Oguri 2012;Richardson et al 2012Richardson et al , 2013). As we argue in §4.1, this "excess" is, in fact, probably close-to-consistent with the amplitude of clustering found for some types of galaxies on small scales.…”

“…Moore et al 1996;Navarro, Frenk & White 1997). A full modeling of this effect will require an in-depth study of the "one-halo" term of the HOD (e.g., Kayo & Oguri 2012;Richardson et al 2012Richardson et al , 2013, which we reserve for future work.…”

The dark matter haloes of moderate luminosity X-ray AGN as determined from weak gravitational lensing and host stellar masses

“…This parameter is sometimes interpreted as an AGN duty cycle (e.g. Martini & Weinberg 2001;Shen et al 2007;Miyaji et al 2011;Richardson et al 2013). In this case, the rising nature of N cen could indicate a varying AGN duty cycle with halo mass.…”

“…While an HOD-type approach may work well for galaxy samples defined by simple luminosity thresholds, it is less clear which parametric form should be adopted for occupation functions when considering AGN-type samples (e.g. Allevato et al 2011;Miyaji et al 2011;Kayo & Oguri 2012;Richardson et al 2013). The AGN duty cycle relative to haloes is unknown, which leads to large uncertainties in both the shape and normalization of the AGN occupation functions.…”

Clustering of intermediate redshift quasars using the final SDSS III-BOSS sample

“…Star formation has been shown to occur most efficiently in halos of similar mass to those occupied by quasars at z < 2 (∼ 10 12 h −1 M⊙), and as measurements of star formation as a function of dark matter halo mass are now being pushed to z > 2 (e.g., Viero et al 2013aViero et al , 2014 it will be intriguing to see whether this trend continues to higher redshift. Finally, models of how quasars populate dark matter halos, for instance, empirical approaches such as the halo occupation distribution framework (HOD; Seljak 2000; Peacock & Smith 2000;Benson et al 2000;Scoccimarro et al 2001;White, Hernquist & Springel 2001;Berlind & Weinberg 2002;Cooray & Sheth 2002, and references therein) have been successful in modeling galaxies at low redshift (e.g., Zheng et al 2005;Zheng, Coil & Zehavi 2007;Zehavi et al 2011), but are now being updated to model quasar clustering at different wavelengths and redshifts (e.g., Degraf et al 2011a;Richardson et al 2012;White et al 2012;Conroy & White 2013;Richardson et al 2013;Veale, White & Conroy 2014;Caplar, Lilly & Trakhtenbrot 2014). It is not clear to what extent such HOD models describe rare and heavily biased quasars at high redshift, and it is critical to push observations to z ∼ 3 as the most sophisticated hydrodynamic simulations being used to interpret the HOD are now being pushed from high redshift down to z < 4 (e.g., DeGraf et al 2012;Di Matteo et al 2012).…”